Evolutionary bound constraint handling for particle swarm optimization

Most global optimization problems have some constraints within their solution space and some of them are related to the boundary limitations. Implementation of these restrictions calls a proper method that is readily usable. In the particle swarm optimization (PSO) algorithm, solutions can easily violate of the bounds limitations; therefore, a bound constraint handling (BCH) method affects the algorithm performance considerably. There are a few studies in the literature about this issue in PSO. This demand is tackled in this study by introducing an effective technique for BCH in the PSO algorithm, called evolutionary boundary constraint handling (EBCH). Several benchmark functions are optimized with the EBCH method, and the results are compared with ten other approaches proposed in the literature including general BCH approaches and PSO specific approaches. The results reveal that, in most cases, the EBCH can considerably improve the performance of the PSO algorithm in comparison with other BCH methods.